统一多步与少步生成模型，训练采样更高效。
Unified Continuous Generative Models
- 提出统一框架，兼容多步与少步生成方法
- 20步达1.30 FID，2步达1.42 FID，性能领先
- 可提升已有模型表现，适合高效生成场景
连续生成模型近年取得显著进展，包括需8-1000步的扩散模型与流匹配，以及仅需1-8步的一致性模型。然而现有工作常将其视为独立范式，导致训练与采样方法分离。本文提出统一框架，实现训练、采样与分析一体化。所提出的UCGM-T/S在ImageNet 256x256上，使用675M扩散变压器，20步即达1.30 FID，2步达1.42 FID；对预训练模型（原250步1.26 FID）应用UCGM-S后，40步提升至1.06 FID。代码已开源。
原文摘要 · Abstract (English)
Recent advances in continuous generative models, including multi-step approaches like diffusion and flow-matching (typically requiring 8-1000 sampling steps) and few-step methods such as consistency models (typically 1-8 steps), have demonstrated impressive generative performance. However, existing work often treats these approaches as distinct paradigms, resulting in separate training and sampling methodologies. We introduce a unified framework for training, sampling, and analyzing these models. Our implementation, the Unified Continuous Generative Models Trainer and Sampler (UCGM-{T,S}), achieves state-of-the-art (SOTA) performance. For example, on ImageNet 256x256 using a 675M diffusion transformer, UCGM-T trains a multi-step model achieving 1.30 FID in 20 steps and a few-step model reaching 1.42 FID in just 2 steps. Additionally, applying UCGM-S to a pre-trained model (previously 1.26 FID at 250 steps) improves performance to 1.06 FID in only 40 steps. Code is available at: https://github.com/LINs-lab/UCGM.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。